Your Launch Post Was Not Your One Shot: Show HN Strategy

Most of the Hacker News posts that took off for big open source projects were not launches. Here is what they were, why nobody saw your post, and a Show HN strategy for your next one.

Your launch post was not your one shot

Researchers once looked at the Hacker News posts that took off for 96 big open source projects. Hacker News is a website where programmers share links and vote on them. Open source means the code is free for anyone to see and use.

Most of those posts were not launches. The first release was one post in six (Borges and Valente, 2019).

Now think about your own post.

You built a small tool at night, next to your day job. One evening you put it on Show HN. That is the page on Hacker News for things people made.

Then you waited. It got 2 points. Nobody wrote a comment.

Or it got one good day. Visitors came, then stopped. By the weekend the project was quiet again.

So you went back to the code. You told yourself nobody wanted it. You will not post it again.

Hold that thought. If the launch was one post in six, what were the other five? And did anyone even see yours?

What you will know by the end

Most guides on Hacker News say the same three things. Post at the right time. Reply to comments. Have a demo.

They leave out what happened to Show HN in the last two years. And they leave out one fact about the posts that take off.

By the end of this piece you will know:

• Whether anyone actually saw your post.
• The hour that changes your odds four times over, for no extra work.
• What Hacker News itself does with good posts that got missed.
• What one good day is worth, and how long it lasts.
• The kind of post that took off most often for projects like yours.

Let us start with a street of shops.

Why nobody saw your post

Picture a street with three shops. In 2022 you could walk the whole street in five minutes and see each one.

Now the same street has twenty shops. The walk still takes five minutes. Most shops get a glance, or nothing.

That is what happened to Show HN.

From 2012 to 2022, Show HN posts were 2 or 3 in every 100 stories on Hacker News. By December 2025 they were more than 12 in 100. In February 2026 they were about 20 in 100 (snubi.net, January 2026).

In plain numbers: about 1,200 Show HN posts a month in early 2023. About 4,800 a month by January 2026 (Arthur Cnops, 17 February 2026). Spread over a month, that is about 160 new projects a day.

The votes did not grow with them. In December 2025 a Show HN post got 9 points on average. Other stories got about 19 (snubi.net).

Here is the number that matters for you. Line up all 41,301 Show HN posts from June 2025 to June 2026 by points. The one in the middle got 2 points and 0 comments. 62 in 100 posts got no comment at all (jonno.nz, 2026).

So 2 points and silence does not mean people said no. It is the normal result of a post that nobody opened. Your project was most likely never seen.

Hacker News noticed too. In March 2026 dang, one of the two people who run the site, wrote: "We're going to at least restrict Show HNs for a while" (Hacker News, March 2026). New accounts that try to post a Show HN now see a notice. It says: "We're temporarily restricting Show HNs because of a massive influx" (news.ycombinator.com/showlim). A massive influx means a flood of new posts.

He also added a line to his guide on 28 March 2026: "Write your text by hand. Don't use an LLM to generate any of it" (Hacker News). An LLM is an AI writing tool.

So the street is crowded, and the door is half shut for newcomers. But the hour you walked onto the street mattered too.

The hour that changes your odds four times over

One person counted 188,085 Show HN posts from 2012 to April 2026 (danfking, 23 April 2026).

He sorted them by the hour they were posted. Then he counted how many reached 50 points, which is a strong result.

Posted on Sunday at 7pm US Eastern time, 10.8 in 100 posts reached 50 points. That is midnight on Monday in UTC, the clock time the internet uses.

Posted on Thursday at 6am UTC, 2.6 in 100 did.

Same kind of post. Four times the odds. The only thing that changed was the clock.

Why? Nobody can say for sure. One guess is that Sunday evening has few new posts and many readers. Think of walking onto that street when it is quiet, and people are looking.

One honest limit. 62 in 100 posts get no comment at any hour. The clock helps a thing people can try and want to talk about. It does not rescue a thing they cannot try. The Show HN rules say it plainly. Make it easy to try, "ideally without barriers such as signups or emails" (Show HN rules).

So say your post does get seen. What does one good day actually give you?

What a good day on Hacker News is really worth

It is fair to say this first. A good day on Hacker News is real, and it is worth having.

Two researchers, Hudson Borges and Marco Tulio Valente, studied the 100 most starred projects on GitHub. GitHub is the website where programmers keep their code. A star is a bookmark. People star a project to say "I like this".

They found 3,019 Hacker News posts about 96 of those projects. They picked the successful ones, the top 10 in 100 by votes. Then they counted stars before and after.

In the three days before a successful post, the project in the middle gained 74 stars. In the three days after, it gained 138 (Borges and Valente, 2019).

The 2026 count found something similar. Each vote up on Show HN brought about 1.4 GitHub stars inside two days. For the top projects, the first day ran at about 509 new stars. The day before, it was less than one (danfking, 2026).

Stars matter because people check them. An earlier study looked at 5,000 projects and asked 115 developers. 3 in 4 said they look at the star count before using a project (Borges and Valente, 2018). A few projects grew very fast. For 73 in 100 of those, the makers said it came from a post on a site like Hacker News.

And Hacker News does give missed posts a second look. Since 2014 the people who run the site have kept a "second chance pool". They read old posts by hand and put good ones back on the front page, the first page people see. dang wrote in 2021 that anyone can point out a missed post by email. His words: "please tell us at hn@ycombinator.com" (Hacker News, 30 April 2021). In January 2026 tomhow, the other person who runs the site, said good projects do get missed. His words: "we can always put things in the second chance pool" (Hacker News, January 2026).

One rule goes with that. The site's own help page says: "Please don't delete and repost the same story" (Hacker News FAQ).

So a hit is worth having. But that is only half of it. Even a hit has a clock on it.

Why even a hit is over in two days, and what took off instead

Go back to the 2026 count of 188,085 posts.

For the top projects, 92 in 100 of the new stars arrived inside 48 hours. By day two the project in the middle was down to 40 stars a day. After a week, almost nothing (danfking, 2026).

Comments move even faster. Across 41,301 posts, half of all comments arrived inside 7.2 hours. 90 in 100 arrived by hour 26 (jonno.nz, 2026).

Big launches and small ones had the same clock. The author put it this way: "Success multiplies how many people show up. It does nothing to change when they leave."

So the quiet after your good day was not a sign of anything. It happens to every post, big or small.

Now back to the question at the start. What were the other five posts in six?

Borges and Valente sorted the successful posts by what they announced. 44.9 in 100 announced a new version of the project. 25.4 in 100 were articles about the project. 10.6 in 100 showed a new feature. Only 16.5 in 100 were the first release (Borges and Valente, 2019).

Most of the posts that took off announced a new version. The project had already launched. It was showing up again with something new.

That is the whole answer. The launch post was one post. The next real version was the next one.

There is one more piece. In the 2018 study, the projects that grew fastest had been updated in the last few days. The fast growing project in the middle had its last code change about half a week before. The slow project in the middle, about three and a half weeks before. Makers of slow projects named "low activity" as the first reason, 54 in 100 of them (Borges and Valente, 2018).

This is a link, not a proven cause. But it points the same way. Posts follow work that is still moving.

So here is what to do with your project this week.

Your Show HN strategy for this week

1. Check whether your post was seen. Open it. If it has 2 points and no comments, it was not read. That is the normal result. It does not mean people said no.
2. Be a person on Hacker News first. Read, comment and vote for a few weeks. Most new accounts cannot post a Show HN right now. The people who run the site said taking part first is reasonable.
3. Make the thing easy to try with no sign up. One link people can click and use in a minute.
4. Write the post by hand. Say what it does, why you built it, and what is not finished yet.
5. Post on Sunday evening US time. That is Monday 00:00 UTC. Set an alarm.
6. Stay at the keyboard for seven hours. Half the comments arrive in that window. Answer every one.
7. If it sinks, do not delete and repost. Someone who thinks it was good can tell hn@ycombinator.com. The people who run the site have said they cannot answer Show HN emails since March 2026. So do not count on it.
8. When the next real version ships, post it as its own Show HN. That is the kind of post that took off most often.
9. Keep the code moving between posts. The projects that kept growing were the ones still being worked on.

What these numbers cannot tell you

The star numbers come from the top 500 or top 5,000 projects on GitHub. A side project with 30 stars is not in those groups. Its numbers may look different.

The 2019 study is about open source projects, and about posts by anyone, not only the maker. A project whose code is private, with only a web page, is a different case.

The "still being worked on" finding is a link. Nobody has tested whether more updates cause more stars.

The second chance pool is small and run by hand. Since March 2026 the people who run the site say they cannot answer Show HN emails.

The 2026 star count stopped at 1,000 stars per project and covers only the top posts. The Show HN counts come from people outside Hacker News, using its public data. Hacker News itself has not published them.

Frequently asked questions about a Show HN launch

What is Show HN?

Show HN is the part of Hacker News where people share something they made that others can try. The rules say it must be something people can play with, not a sign up page.

Why can I not post a Show HN with a new account?

In spring 2026 Hacker News limited Show HN posts from accounts with little history. There was a flood of posts from people new to the site. Taking part on the site for a while comes first.

Can I post my project again after a new version?

Yes. A real new version is a new post. In one study of 96 big projects, 44.9 in 100 of the posts that took off announced a new version. Do not delete and repost the same post.

Does the posting hour really matter?

In a count of 188,085 posts, 10.8 in 100 posted on Sunday 7pm US Eastern reached 50 points. On Thursday 6am UTC, 2.6 in 100 did. It helps a post people want to open. It does not fix one they do not.

One question for you

Did you ever post your project a second time, after a new version?

What happened the second time?